github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2026-06-09 10:17:00 +00:00
Code Issues Packages Projects Releases Wiki Activity
16,096 Commits 1,164 Branches 1,270 Tags
3b999176c84c1236e1f6fbaa5194782360e82264
Commit Graph

364 Commits

Author SHA1 Message Date
Mason Daugherty
31eeb50ce0 chore: drop UP045 (#33362) 
Python 3.9 EOL
 2025-10-08 21:17:53 -04:00
Mason Daugherty
d13823043d style: monorepo pass for refs (#33359) 
* Delete some double backticks previously used by Sphinx (not done
everywhere yet)
* Fix some code blocks / dropdowns

Ignoring CLI CI for now
 2025-10-08 18:41:39 -04:00
Mason Daugherty
6b9b177b89 chore(openai): base.py ref pass (#33355) 2025-10-08 16:08:52 -04:00
ccurme
de48e102c4 fix(core,openai,anthropic): delegate to core implementation on invoke when streaming=True (#33308) 2025-10-06 15:54:55 -04:00
ccurme
95a451ef2c fix(openai): disable stream_usage in chat completions if OPENAI_BASE_URL is set (#33298) 
This env var is used internally by the OpenAI client.
 2025-10-06 10:14:43 -04:00
ccurme
c8636a626a chore(openai): (v1) fix sort order of mcp call keys (#33295) 2025-10-06 09:29:41 -04:00
ccurme
4e50ec4b98 feat(openai): enable stream_usage when using default base URL and client (#33205) 2025-10-06 08:56:38 -04:00
Mason Daugherty
8e7cd85431 style: drop target-version = "py39" for OpenAI, Anthropic, HuggingFace (#33287) 2025-10-06 03:29:34 +00:00
Mason Daugherty
ae5b105d11 docs: v1 docs updates (#33173) 
Co-authored-by: Mohammad Mohtashim <45242107+keenborder786@users.noreply.github.com>
Co-authored-by: Caspar Broekhuizen <caspar@langchain.dev>
Co-authored-by: ccurme <chester.curme@gmail.com>
Co-authored-by: Christophe Bornet <cbornet@hotmail.com>
Co-authored-by: Eugene Yurtsev <eyurtsev@gmail.com>
Co-authored-by: Sadra Barikbin <sadraqazvin1@yahoo.com>
Co-authored-by: Vadym Barda <vadim.barda@gmail.com>
 2025-10-02 18:46:26 -04:00
Mason Daugherty
eaa6dcce9e release: v1.0.0 (#32567) 
Co-authored-by: Mohammad Mohtashim <45242107+keenborder786@users.noreply.github.com>
Co-authored-by: Caspar Broekhuizen <caspar@langchain.dev>
Co-authored-by: ccurme <chester.curme@gmail.com>
Co-authored-by: Christophe Bornet <cbornet@hotmail.com>
Co-authored-by: Eugene Yurtsev <eyurtsev@gmail.com>
Co-authored-by: Sadra Barikbin <sadraqazvin1@yahoo.com>
Co-authored-by: Vadym Barda <vadim.barda@gmail.com>
 2025-10-02 10:49:42 -04:00
ccurme
64141072a3 feat(openai): support openai sdk 2.0 (#33168) 2025-09-30 16:34:00 -04:00
Mason Daugherty
986302322f docs: more standardization (#33124) 2025-09-25 20:46:20 -04:00
Mason Daugherty
5bea28393d docs: standardize .. code-block directive usage (#33122) 
and fix typos
 2025-09-25 16:49:56 -04:00
Mason Daugherty
9f6431924f feat(openai): add max_tokens to AzureChatOpenAI (#32959) 
Fixes #32949

This pattern is [present in
`ChatOpenAI`](https://github.com/langchain-ai/langchain/blob/master/libs/partners/openai/langchain_openai/chat_models/base.py#L2821)
but wasn't carried over to Azure.


[CI](https://github.com/langchain-ai/langchain/actions/runs/17741751797/job/50417180998)
 2025-09-15 14:09:20 -04:00
Matthew Lapointe
b1f08467cd feat(core): allow overriding ls_model_name from kwargs (#32541) 2025-09-11 16:18:06 -04:00
Aasish
9c7d262ff4 fix(openai): update AzureOpenAIEmbeddings validation logic for openai_api_base (#31782) 2025-09-10 14:53:30 -04:00
Mason Daugherty
4c6af2d1b2 fix(openai): structured output (#32551) 2025-09-09 11:37:50 -04:00
Sadiq Khan
228fbac3a6 fix(openai): handle AIMessages without response_id in _get_last_messages (#32824) 2025-09-08 10:12:50 -04:00
JunHyungKang
6ea06ca972 fix(openai): Fix Azure OpenAI Responses API model field issue (#32649) 2025-09-08 10:08:35 -04:00
ccurme
5b0a55ad35 chore(openai): apply formatting changes to AzureChatOpenAI (#32848) 2025-09-08 09:54:20 -04:00
Shahroz Ahmad
4828a85ab0 feat(core): add web_search in OpenAI tools list (#32738) 2025-09-02 21:57:25 +00:00
Ravirajsingh Sodha
b42dac5fe6 docs: standardize OllamaLLM and BaseOpenAI docstrings (#32758) 
- Add comprehensive docstring following LangChain standards
- Include Setup, Key init args, Instantiate, Invoke, Stream, and Async
sections
- Provide detailed parameter descriptions and code examples
- Fix linting issues for code formatting compliance

Contributes to #24803

---------

Co-authored-by: Mason Daugherty <github@mdrxy.com>
 2025-08-31 17:45:56 -05:00
Jacob Lee
1459d4f4ce fix(openai): Always add raw response object to OpenAI client errors for invoke (#32655) 2025-08-26 09:59:25 -04:00
Alex Naidis
21f7a9a9e5 fix(openai): allow temperature parameter for gpt-5-chat models (#32624) 2025-08-21 16:40:10 -04:00
sa411022
61bc1bf9cc fix(openai): construct responses api input (#32557) 2025-08-21 15:56:29 -04:00
Shahrukh Shaik
4ba222148d fix(openai): Chat Message Annotations defaults to [ ] if not list or None (#32614) 2025-08-21 15:30:12 -04:00
Mason Daugherty
0024dffa68 feat(openai): officially support verbosity (#32470) 2025-08-12 16:00:30 +00:00
Mason Daugherty
c31236264e chore: formatting across codebase (#32466) 2025-08-08 10:20:10 -04:00
Mason Daugherty
00244122bd feat(openai): minimal and verbosity (#32455) 2025-08-08 02:24:21 +00:00
Michael Matloka
5036bd7adb fix(openai): don't crash get_num_tokens_from_messages on gpt-5 (#32451) 2025-08-07 16:33:19 -04:00
ccurme
ec2b34a02d feat(openai): custom tools (#32449) 2025-08-07 16:30:01 -04:00
Mason Daugherty
145d38f7dd test(openai): add tests for prompt_cache_key parameter and update docs (#32363) 
Introduce tests to validate the behavior and inclusion of the
`prompt_cache_key` parameter in request payloads for the `ChatOpenAI`
model.
 2025-08-07 15:29:47 -04:00
ccurme
68c70da33e fix(openai): add in output_text (#32450) 
This property was deleted in `openai==1.99.2`.
 2025-08-07 15:23:56 -04:00
Mason Daugherty
96cbd90cba fix: formatting issues in docstrings (#32265) 
Ensures proper reStructuredText formatting by adding the required blank
line before closing docstring quotes, which resolves the "Block quote
ends without a blank line; unexpected unindent" warning.
 2025-07-27 23:37:47 -04:00
niceg
0d6f915442 fix: LLM mimicking Unicode responses due to forced Unicode conversion of non-ASCII characters. (#32222) 
fix: Fix LLM mimicking Unicode responses due to forced Unicode
conversion of non-ASCII characters.

- **Description:** This PR fixes an issue where the LLM would mimic
Unicode responses due to forced Unicode conversion of non-ASCII
characters in tool calls. The fix involves disabling the `ensure_ascii`
flag in `json.dumps()` when converting tool calls to OpenAI format.
- **Issue:** Fixes ↓↓↓
input：
```json
{'role': 'assistant', 'tool_calls': [{'type': 'function', 'id': 'call_nv9trcehdpihr21zj9po19vq', 'function': {'name': 'create_customer', 'arguments': '{"customer_name": "你好啊集团"}'}}]}
```
output:
```json
{'role': 'assistant', 'tool_calls': [{'type': 'function', 'id': 'call_nv9trcehdpihr21zj9po19vq', 'function': {'name': 'create_customer', 'arguments': '{"customer_name": "\\u4f60\\u597d\\u554a\\u96c6\\u56e2"}'}}]}
```
then:
llm will mimic outputting unicode. Unicode's vast number of symbols can
lengthen LLM responses, leading to slower performance.
<img width="686" height="277" alt="image"
src="https://github.com/user-attachments/assets/28f3b007-3964-4455-bee2-68f86ac1906d"
/>

---------

Co-authored-by: Mason Daugherty <github@mdrxy.com>
Co-authored-by: Mason Daugherty <mason@langchain.dev>
 2025-07-24 17:01:31 -04:00
Mason Daugherty
d53ebf367e fix(docs): capitalization, codeblock formatting, and hyperlinks, note blocks (#32235) 
widespread cleanup attempt
 2025-07-24 16:55:04 -04:00
Copilot
54542b9385 docs(openai): add comprehensive documentation and examples for extra_body + others (#32149) 
This PR addresses the common issue where users struggle to pass custom
parameters to OpenAI-compatible APIs like LM Studio, vLLM, and others.
The problem occurs when users try to use `model_kwargs` for custom
parameters, which causes API errors.

## Problem

Users attempting to pass custom parameters (like LM Studio's `ttl`
parameter) were getting errors:

```python
# ❌ This approach fails
llm = ChatOpenAI(
    base_url="http://localhost:1234/v1",
    model="mlx-community/QwQ-32B-4bit",
    model_kwargs={"ttl": 5}  # Causes TypeError: unexpected keyword argument 'ttl'
)
```

## Solution

The `extra_body` parameter is the correct way to pass custom parameters
to OpenAI-compatible APIs:

```python
# ✅ This approach works correctly
llm = ChatOpenAI(
    base_url="http://localhost:1234/v1",
    model="mlx-community/QwQ-32B-4bit",
    extra_body={"ttl": 5}  # Custom parameters go in extra_body
)
```

## Changes Made

1. **Enhanced Documentation**: Updated the `extra_body` parameter
docstring with comprehensive examples for LM Studio, vLLM, and other
providers

2. **Added Documentation Section**: Created a new "OpenAI-compatible
APIs" section in the main class docstring with practical examples

3. **Unit Tests**: Added tests to verify `extra_body` functionality
works correctly:
- `test_extra_body_parameter()`: Verifies custom parameters are included
in request payload
- `test_extra_body_with_model_kwargs()`: Ensures `extra_body` and
`model_kwargs` work together

4. **Clear Guidance**: Documented when to use `extra_body` vs
`model_kwargs`

## Examples Added

**LM Studio with TTL (auto-eviction):**
```python
ChatOpenAI(
    base_url="http://localhost:1234/v1",
    api_key="lm-studio",
    model="mlx-community/QwQ-32B-4bit",
    extra_body={"ttl": 300}  # Auto-evict after 5 minutes
)
```

**vLLM with custom sampling:**
```python
ChatOpenAI(
    base_url="http://localhost:8000/v1",
    api_key="EMPTY",
    model="meta-llama/Llama-2-7b-chat-hf",
    extra_body={
        "use_beam_search": True,
        "best_of": 4
    }
)
```

## Why This Works

- `model_kwargs` parameters are passed directly to the OpenAI client's
`create()` method, causing errors for non-standard parameters
- `extra_body` parameters are included in the HTTP request body, which
is exactly what OpenAI-compatible APIs expect for custom parameters

Fixes #32115.

<!-- START COPILOT CODING AGENT TIPS -->
---

💬 Share your feedback on Copilot coding agent for the chance to win a
$200 gift card! Click
[here](https://survey.alchemer.com/s3/8343779/Copilot-Coding-agent) to
start the survey.

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: mdrxy <61371264+mdrxy@users.noreply.github.com>
Co-authored-by: Mason Daugherty <github@mdrxy.com>
Co-authored-by: Mason Daugherty <mason@langchain.dev>
 2025-07-24 16:43:16 -04:00
ccurme
de13f6ae4f fix(openai): support acknowledged safety checks in computer use (#31984) 2025-07-14 07:33:37 -03:00
ccurme
612ccf847a chore: [openai] bump sdk (#31958) 2025-07-10 15:53:41 -04:00
Mason Daugherty
6594eb8cc1 docs(xai): update for Grok 4 (#31953) 2025-07-10 11:06:37 -04:00
Mason Daugherty
33c9bf1adc langchain-openai[patch]: Add ruff bandit rules to linter (#31788) 2025-06-30 14:01:32 -04:00
Andrew Jaeger
0189c50570 openai[fix]: Correctly set usage metadata for OpenAI Responses API (#31756) 2025-06-27 15:35:14 +00:00
ccurme
e8e89b0b82 docs: updates from langchain-openai 0.3.26 (#31764) 2025-06-27 11:27:25 -04:00
ccurme
88d5f3edcc openai[patch]: allow specification of output format for Responses API (#31686) 2025-06-26 13:41:43 -04:00
ccurme
84500704ab openai[patch]: fix bug where function call IDs were not populated (#31735) 
(optional) IDs were getting dropped in some cases.
 2025-06-25 19:08:27 +00:00
ccurme
0bf223d6cf openai[patch]: add attribute to always use previous_response_id (#31734) 2025-06-25 19:01:43 +00:00
joshy-deshaw
8a0782c46c openai[patch]: fix dropping response headers while streaming / Azure (#31580) 2025-06-23 17:59:58 -04:00
ccurme
b268ab6a28 openai[patch]: fix client caching when request_timeout is specified via httpx.Timeout (#31698) 
Resolves https://github.com/langchain-ai/langchain/issues/31697
 2025-06-23 14:37:49 +00:00
Li-Kuang Chen
4ee6112161 openai[patch]: Improve error message when response type is malformed (#31619) 2025-06-21 14:15:21 -04:00
ccurme
e2a0ff07fd openai[patch]: include 'type' key internally when streaming reasoning blocks (#31661) 
Covered by existing tests.

Will make it easier to process streamed reasoning blocks.
 2025-06-18 15:01:54 -04:00